The cells of solid tumors often have high levels of chromosome rearrangements (deletions, duplications, and translocations), and abnormal numbers of chromosomes (aneuploidy). The mechanisms giving rise to this instability are not understood. Our goal is to investigate the mechanisms responsible for genetic instability in tumor cells by studying the genetic regulation of instability in the yeast Saccharomyces cerevisiae. We have developed methods that allow us to locate the positions of chromosome rearrangements throughout the yeast genome using oligonucleotide-containing microarrays. Since chromosome rearrangements are often generated by homologous recombination between repeated genes, several aspects of mitotic recombination will be investigated. One important issue is the control of the bias of DNA repair events toward utilization of the sister- chromatid rather than the homolog as a substrate for repair. A second issue is the mechanism by which elevated levels of transcription stimulate recombination. We have developed a genetic system that allows investigation of both of these issues. Recombination is elevated in yeast strains that have defects in DNA replication or problems with chromosome condensation. Using DNA microarrays, we will investigate the breakpoints for chromosome rearrangements in mutant strains that have low levels of DNA polymerase delta (one of the replicative DNA polymerases) or defects in chromosome condensation. In particular, the relationship between regions with slow-moving replication forks and chromosome rearrangement breakpoints will be examined. It has recently been shown that expression of the Tus protein of E. coli results in a stalled DNA replication fork at a Ter site (also derived from E. coli) inserted into a yeast chromosome. We will test whether this stalled replication fork is a hotspot for mitotic recombination in yeast. Tumor cells often exhibit uniparental disomy (UPD), loss of one homolog and duplication of the other. We have developed a system for detection of UPD in yeast, and shown that strains that lack the Top2 DNA topoisomerase have very elevated levels of UPD. We hypothesize that certain types of UPD events reflect the entanglement of homologs, and we propose experiments to test this hypothesis. In mammalian cells, internally-located telomeric sequences (ITSs) are frequently observed at the breakpoints of tumor-associated chromosome rearrangements. We have developed methods of detecting this type of instability in yeast, and will examine which proteins influence the rate of ITS-associated rearrangements. Lastly, most mutant hunts are performed in yeast by treating strains with a mutagen that increases the rate of genomic point mutations. We will use mutant yeast strains with very high levels of chromosome rearrangements to search for variants with novel phenotypes such as the ability to grow at elevated temperatures or to grow faster than wild-type strains.